NUNATAK AND CASCADING GLACIERS 135 the position of its 1899 terminus the hill and valley surface consists of patches of bare, grooved, and polished rocks which frost action has not had time to roughen greatly, alternating with gravel and moraine areas in which the vegetation is young. It cannot have been many decades since this tongue pushed out far enough to have united with the tidal arm of Nunatak Glacier. It is possible, though not probable, that it did so in 1891 and it is unfortunate that Russell did not obtain near enough views of this glacier for us to know its position with certainty. The Canadian Boundary survey photographs show the hill as a semi-nunatak in 1895, with the land tongue occupying the head of the gorge. The first photograph showing its position from near at hand was taken by Gilbert in 1899, but we found in 1909 and 1910 some evidence for the preceding year. In 1898 some 300 prospectors passed over the Nunatak Glacier highway to the Alsek valley, and some returned by this route. On their return they abandoned their outfits, presumably as soon as they were useless and at the point farthest down the glacier over which they could draw their sledges. Roughly, with an error of only a few yards, the places where we found these sledges and other relics would represent the position of the ice front in 1898. We found a half dozen sledges, several rusty Yukon stoves, some ice creepers, spiked boots, etc. (PI. LXII, B), at a point a little less than a half mile west of the 1909 and 1910 ice fronts and just south of where Gilbert's photographs show the terminus to have been in 1899. Aside from the improbability that men would draw sledges over bare rock (in the fall) and up hill and beyond the glacier terminus, the fact that the outfits were not being taken beyond where the ice ended in 1898 is established by the presence of an impassable stream gorge between the positions of the sledges, which were in two groups, and the present glacier terminus. It is, therefore, concluded that the ice ended in 1898 about where the sledges were found and that the retreat of the land distributary between 1898 and 1909 has been a little less than a half mile. In each of our five years of observation here, 1905, 1906, 1909, 1910 and 1918, we carefully observed the land tongue pf Nunatak Glacier but without finding any evidence of advance. Being a glacier distributary still connected with an active glacier, but itself stagnant, it would be very quick to show response to even moderate advance. The fact that we have not detected any advance, together with the evidence that there was pronounced recession between 1899 and 1905, is accepted by us as definite proof that as yet there has been no advance as a result of the earthquakes of 1899. In 1910 and 1913, the junior author observed a continuation of retreat of the land arm of Nunatak Glacier which was then not essentially different from the 1909 condition except in the thinness of the terminal feather edge. The moraine to the right of it seemed to stand slightly higher in relief in 1910 than the year before, as comparative photographs showed. EIQ. 10. NUNATAK GLACIHB IN 1905, SHOWING GOBOEJ NEAH LAND-ENDING ABM.